“Normally for glaucoma, we’re monitoring pressure two to four times a year with patients coming in every three to six months for a pressure check in the clinic, and it’s just a single spot reading. At home with this device, you get multiple readings over continuous days, so it’s a more accurate representation of their pressure in real life. You can catch spikes in pressure that you might miss from that one in-office check.”Leo Seibold, MD, Associate Professor of Ophthalmology, University of Colorado School of Medicine
Glaucoma is an age-related chronic optic neuropathy and the leading cause of irreversible blindness worldwide. Predictions report that in 2040 the number of people with glaucoma worldwide will reach almost 112 million, disproportionally affecting Asian and African countries. A significant challenge in trying to tackle glaucoma-related blindness is to identify those with this eye disease before they become symptomatic.
Glaucoma is diagnosed clinically by detecting the characteristic changes of the optic disc and can be confirmed with a corresponding visual field (VF) defect. However, the clinical diagnosis of glaucoma is subjective and relies on the examiner’s experience. A particular challenge for detecting glaucoma is the wide variation of optic disc structure and size in the population. Therefore, there is an urgent need for novel and accurate techniques to detect glaucoma that can be used in different settings.
Fortunately, we are witnessing new practical applications of artificial intelligence (AI), including advances in complex artificial neural networks (ANNs) and constant innovations in daily-used devices such as smartphones, laptops, electronic tablets, wireless communications, etc. Here is a quick review of some new technologies that improve glaucoma diagnosis and treatment.
There are several well-recognized risk factors for glaucoma, including elevated IOP, older age, ethnicity, and family history of glaucoma. Of all risk factors, the level of IOP is the most important one and is the only one that can be modified with treatment. The risk of developing glaucoma rises with increasing IOP. This is supported by the fact that those patients presenting with advanced disease at diagnosis are more likely to have higher IOP. Higher IOP is also a risk factor for disease progression.
Self-monitoring of chronic diseases, such as monitoring blood glucose levels for diabetes mellitus and measuring blood pressure in hypertensive patients, can be effective for informing patients when their current treatment is not sufficiently effective, thereby giving them a sense of control over their condition along with improved disease control. The ability to measure IOP at home throughout different periods of the day will enable physicians to identify potential failures in treatment and possibly improve the prognosis for the patient. The self-monitoring approach could bring significant patient benefits, such as symptom management and improved quality of life.
Peak IOP and IOP fluctuations have been identified as risk factors for the development and progression of glaucoma. Diurnal fluctuations of IOP are highly unlikely to be observed during clinical visits. The Icare home tonometer is based on the principle of rebound tonometry. No topical anesthetic is required, and there is minimal risk of corneal injury. It may be an option for some glaucoma patients to monitor their IOP better at home and aid in managing their condition.
New technologies, such as sensors and wireless devices, present promising tools that enable continuous monitoring of IOP. One of the first platforms used was the SENSIMED Triggerfish contact lens sensor (Sensimed AG, Lausanne, Switzerland). This soft silicone contact lens is provided with a circumferential sensor that consists of 2 platinum-titanium strain gauges designed to measure changes in the radius of curvature of the cornea. A microprocessor transmits an output signal to a wireless antenna on the periocular surface. A cable wire transfers the data to an external portable recorder. The SENSIMED Triggerfish is noninvasive, but the data’s clinical value is uncertain.
Assessment of the damaged optic nerve and retinal nerve fiber layer is a crucial method to detect glaucoma. Stereoscopic fundus examination by an expert clinician is considered standard practice. Automated imaging systems, particularly optical coherence tomography (OCT), are now widely used for glaucoma diagnosis. However, OCT examinations involve high costs, which may not be affordable by some healthcare providers. The advent of digital photography has made it easier and cheaper to acquire and process optic disc images. Nonmydriatic stereoscopic cameras are particularly useful for evaluating the optic disc and helping detect disease progression. Using image processing, relevant features, such as the optic disc and blood vessels, can be analyzed and provide helpful information. Advanced mobile phone technologies enable remote health care delivery and have been proposed as useful tools in glaucoma detection. Newer smartphone devices have high-powered computational functions, cameras, image processing, and communication capabilities, and they have been developed as an inexpensive retinal photography tools.
One example of a successful combination of fundal imaging and smartphone technology is the Portable Eye Examination Kit (PEEK) Retina (Peek Vision, London, United Kingdom), a smartphone camera adapter developed by Bastawrous et al. This portable mobile phone retinal imaging system is low-cost and easy to use for minimally trained users. The PEEK Retina can capture excellent-quality optic disc and retinal images and has recently been validated for optic nerve imaging in population-based studies in the developing world. Multiple other adapters have been developed to convert modern smartphone cameras into fundus and anterior segment cameras, thus reducing their costs and increasing potential adoption. These include the D-Eye system from Italy (D-Eye S.r.l, Padova, Italy) and the Ocular CellScope (Cellscope Inc, San Francisco, CA) from the United States.
Full-threshold automated perimetry (e.g., Humphrey, Octopus) is routinely used to confirm the diagnosis of glaucoma and monitor disease progression. Alternatives to clinic-based standard automated perimetry have been developed, often to facilitate home monitoring. Smartphones and tablets have been shown to be suited for vision testing, and they can communicate with other wireless platforms sending data automatically for analysis.
Although there are substantial challenges associated with home tonometry, including the calibration of monitors (e.g., securing a constant illumination intensity and size of the stimuli), determining the reliability of the test, and maintaining participant concentration, at the moment it is possible to perform VF tests using a tablet-based procedure. In addition to the role of detecting glaucoma, new perimetry technologies using tablets or portable computers have been designed to facilitate patients performing unsupervised perimetric tests as part of a home-monitoring program. Similar strategies have been tried successfully in patients with age-related macular degeneration and with promising results.
The glaucoma screening application called Visual Fields Easy (VFE) is available for the iPad and can be downloaded for free. The VFE showed promising capacity compared with diagnosis based on a Humphrey 24-2 SITA-Standard outcome. The VFE test evaluates 96 test locations (24 per VF quadrant) across the central 30-degree radius in 3 minutes on average, and the testing distance is 33 cm. The test results can be directly printed or emailed using Wi-Fi from the iPad. VFE worked well for identifying moderate-to-severe disease.
Cambridge Consultants has developed a smartphone glaucoma screening technology called Viewi that could be used in a clinic setting or from patients’ homes. The Viewi system includes a smartphone app, a Bluetooth finger button, and a headset that holds smartphones. The smartphone is slid into the viewer, the app is started, and the patient can hold the button in 1 hand, pushing when a light flash is seen. The app runs a suprathreshold static perimetry test. The test results are displayed in an intuitive format on the smartphone and can be shared instantly with healthcare professionals.
Using ANN and other AI strategies as glaucoma diagnostic tools are becoming a reality. Several techniques have been used to automate the glaucoma detection process. Moorfields Eye Hospital’s collaboration with Google (Alphabet Inc, Mountain View, CA) DeepMind aims to create a general-purpose AI algorithm that can look at OCT scans and diagnose age-related macular degeneration and diabetic retinopathy. VISULYTIX recently introduced PEGASUS (Visulytix Ltd), a new deep learning-based AI technology. It consists of an inexpensive smartphone clip-on optic nerve scanner that looks at several features within fundus images and provides information about the health of the optic disc. PEGASUS can be used not only for glaucoma detection but also for diabetic retinopathy and macular diseases.
Although the number of AI successes in glaucoma will likely grow in the near future, some promising applications are far from their clinical validation. Deep learning AI algorithms might allow the assimilation of multiple composite test results obtained at one screening visit to optimize diagnostic performance at a low cost.
There have been an increasing number of genetic studies on glaucoma recently, mainly due to a substantial reduction in the cost of high-throughput genome-wide genotyping platforms. In glaucoma, it seems that genetic testing may be helpful in some circumstances, such as screening of family members in OAG affecting young people with an apparent autosomal dominant inheritance. It has been proposed that genetic testing may help predict conversion from ocular hypertension to glaucoma or help predict disease progression, but the validation of these observations is required. Several international collaborations are currently ongoing, trying to understand the genetic components of glaucoma better. However, at the moment, genetic testing of broader populations is not justified.
New technologies for glaucoma detection have emerged, but their potential value and adoption greatly depend on the setting and the potential use of the test. Possible adoption of teleglaucoma and portable tablet or smartphone-based technologies is expected to facilitate glaucoma detection and management in remote and underserved populations.